La incorporación de la inteligencia artificial en las políticas públicas de Brasil y Colombia desde una perspectiva comparada

Abstract

This article examines how artificial intelligence has been incorporated into the formulation and implementation of public policies in Brazil and Colombia, following a path that begins with the technical origins of the technology, continues with the states that first brought it into government management and ends with a contrast between the two South American countries. The approach is qualitative, reflective and interpretive, and relies on a documentary design whose declared methodological axis is policy transfer and diffusion, complemented by legal transplant theory for the examination of the European regulatory reference. The corpus is a purposive, non exhaustive sample of eighty-seven items including norms, case law, policy documents, reports from multilateral organizations and peer reviewed academic literature, selected according to explicit criteria of thematic relevance, institutional rank and source verifiability, with publication dates between 1950 and August 2026. The findings show that Brazil built its institutional framework on a previously consolidated normative base, has considerably greater fiscal capacity and has brought the technology into almost all of its courts and oversight bodies, while Colombia formalized its specific national artificial intelligence policy four years later, with a more modest budget allocation and an adoption concentrated in the implementation stage of policy. In both countries the evidence indicates that the technology spreads faster in service operations than in the design of public decisions, that binding regulation advances slowly compared with the pace of adoption and that safeguards against algorithmic opacity still depend on instruments of a merely guiding nature whose legal enforceability is limited. The work concludes that the gap between the two cases lies in accumulated institutional capacity rather than in declared political will.